OBSERVER VARIABILITY IN PINNIPED COUNTS: GROUND-BASED ENUMERATION OF WALRUSES AT HAUL-OUT SITES

Pinnipeds are often monitored by counting individuals at haul-out sites, but the often large numbers of densely packed individuals at these sites are difficult to enumerate accurately. Errors in enumeration can induce bias and reduce precision in estimates of population size and trend. We used data from paired observers monitoring walrus haul-outs in Bristol Bay, Alaska, to quantify observer variability and assess its relative importance. The probability of a pair of observers making identical counts was < 0.1 for walrus groups with >50 individuals. Mean count differences ranged up to 25% for the largest counts, depending on beach and observers. In at least some cases, there was a clear tendency for counts of one observer to be consistently greater than counts of the other observer in a pair, indicating that counts of at least one of the observers were biased. These results suggest that efforts to improve accuracy of counts will be worthwhile. However, we also found that variation among observers was relatively small compared to variation among visits to a beach so that efforts to account for other sources of variation will be more important.     

Patterns of Dolphin Bycatch in a North-Western Australian Trawl Fishery

The bycatch of small cetaceans in commercial fisheries is a global wildlife management problem. We used data from skippers'' logbooks and independent observers to assess common bottlenose dolphin (Tursiops truncatus) bycatch patterns between 2003 and 2009 in the Pilbara Trawl Fishery, Western Australia. Both datasets indicated that dolphins were caught in all fishery areas, across all depths and throughout the year. Over the entire datasets, observer reported bycatch rates (n = 52 dolphins in 4,124 trawls, or 12.6 dolphins/1,000 trawls) were ca. double those reported by skippers (n = 180 dolphins in 27,904 trawls, or 6.5 dolphins/1,000 trawls). Generalised Linear Models based on observer data, which better explained the variation in dolphin bycatch, indicated that the most significant predictors of dolphin catch were: (1) vessel - one trawl vessel caught significantly more dolphins than three others assessed; (2) time of day – the lowest dolphin bycatch rates were between 00:00 and 05:59; and (3) whether nets included bycatch reduction devices (BRDs) - the rate was reduced by ca. 45%, from 18.8 to 10.3 dolphins/1,000 trawls, after their introduction. These results indicated that differences among vessels (or skippers'' trawling techniques) and dolphin behavior (a diurnal pattern) influenced the rates of dolphin capture; and that spatial or seasonal adjustments to trawling effort would be unlikely to significantly reduce dolphin bycatch. Recent skipper''s logbook data show that dolphin bycatch rates have not declined since those reported in 2006, when BRDs were introduced across the fishery. Modified BRDs, with top-opening escape hatches from which dolphins might escape to the surface, may be a more effective means of further reducing dolphin bycatch. The vulnerability of this dolphin population to trawling-related mortality cannot be assessed in the absence of an ongoing observer program and without information on trawler-associated dolphin community size, broader dolphin population size and connectivity with adjacent populations.     

AERIAL SURVEYS FOR COASTAL DOLPHINS: ABUNDANCE OF HECTOR'S DOLPHINS OFF THE SOUTH ISLAND WEST COAST, NEW ZEALAND

Line-transect surveys of abundance depend critically on the estimation of detection probability, which includes corrections for availability and visibility. In this aerial line-transect survey for Hector's dolphin ( Cephalorhynchus hectori ) we recorded dive times from a helicopter to estimate the proportion of time that dolphins could be seen at or near the water surface and were therefore "available" to be counted. The proportion of available sightings that were counted on transect lines was estimated by comparing sightings by two independent teams of two observers in the fixed-wing plane. The survey covered the area between Farewell Spit and Milford Sound, off the west coast of the South Island of New Zealand. Survey effort was stratified according to existing distribution data. A total of 142 separate sightings was made on 1,355 km of trackline. Average availability for fifty Hector's dolphin groups observed from the helicopter (161 dive/surface cycles) was 46.3% (CV = 4.2%). Data from the two independent observer teams indicated that 96.2% (CV = 2.3%) of the dolphin groups that were at the surface and on the trackline were seen. The abundance estimate for the South Island west coast, corrected for visibility and availability, was 5,388 Hector's dolphins (CV = 20.6%). The total population estimate for South Island Hector's dolphins is 7,270 (CV = 16.2%).     

ABUNDANCE OF SOUTHERN CALIFORNIA COASTAL BOTTLENOSE DOLPHINS ESTIMATED FROM TANDEM AERIAL SURVEYS

We describe a tandem aerial survey method for bottlenose dolphins ( Tursiops truncatus ) that uses two aircraft and independent observer teams to conduct consecutive surveys of the same coastal strip one hour apart. Alternatively, one aircraft with one observer team surveys the same coastal strip twice over several hours. Using mark-recapture analysis, we corrected survey counts for visibility bias resulting from missing dolphin groups at the surface and submerged groups. Dolphin groups were considered "recaptured" when we determined that both observer teams had detected the same group. This tandem method is highly useful for estimating abundance (and visibility bias) for species where population closure may be assumed between flights. We assumed population closure between flights and matched groups using geographic location, group size, and expected travel rates. We derive a new variance estimator of population size which incorporates group-size variability commonly encounteted in cetacean surveys. From six tandem surveys conducted from 1991 to 1994, we estimated the abundance of southern California coastal bottlenose dolphins to be between 78 (95% CI 60-102) and 271 (240-306) animals, with an average of 140 (128-154). Variability in abundance estimates is likely due to seasonal and interannual movement of animals along the California and Baja California coast. Abundance estimates from tandem surveys averaged 53% higher than dolphin counts obtained from individual survey flights, demonstrating the importance of correcting for visibility bias.     

Computer‐aided calibration for visual estimation of vegetation cover

Question: What precision and accuracy of visual cover estimations can be achieved after repeated calibration with images of vegetation in which the true cover is known, and what factors influence the results? Methods: Digital images were created, in which the true cover of vegetation was digitally calculated. Fifteen observers made repeated estimates with immediate feedback on the true cover. The effects on precision and accuracy through time were evaluated with repeated proficiency tests. In a field trial, cover estimates, before and after calibration, were compared with point frequency data. Results: Even a short time of calibration greatly improves precision and accuracy of the estimates, and can also reduce the influence of different backgrounds, aggregation patterns and experience. Experienced observers had a stronger tendency to underestimate the cover of narrow‐leaved grasses before calibration. The field trial showed positive effects of computer‐based calibration on precision, in that it led to considerably less between‐observer variation for one of the two species groups. Conclusions: Computer‐aided calibration of vegetation cover estimation is simple, self‐explanatory and time‐efficient, and might possibly reduce biases and drifts in estimate levels over time. Such calibration can also reduce between‐observer variation in field estimates, at least for some species. However, the effects of calibration on estimations in the field must be further evaluated, especially for multilayered vegetation.     

Lacaziosis and lacaziosis-like prevalence among wild, common bottlenose dolphins Tursiops truncatus from the west coast of Florida, USA

Lacaziosis (lobomycosis; Lacazia loboi) is a fungal skin disease that naturally occurs only in humans and dolphins. The first reported case of lacaziosis in a bottlenose dolphin Tursiops truncatus occurred in 1970 in Sarasota Bay, Florida, USA, and subsequent photo-ID monitoring of the Sarasota Bay dolphin population has revealed persistence of the disease. The objectives of this study were to estimate lacaziosis prevalence (P) in 2 bottlenose dolphin populations on the west coast of Florida (Sarasota Bay and Charlotte Harbor) and compare disease occurrence to other published estimates of lacaziosis in dolphin populations across the globe. Historic photographic records of dolphins captured and released for health assessment purposes (Sarasota Bay) and photo-ID studies (Charlotte Harbor) were screened for evidence of lesions consistent with lacaziosis. Health assessment data revealed a prevalence of lacaziosis in the Sarasota Bay bottlenose dolphin population between 2 and 3%, and analyses of photo-ID data provided a lacaziosis-like prevalence estimate of 2% for Charlotte Harbor dolphins. With the exception of lacaziosis prevalence estimates for dolphins inhabiting the Indian River Lagoon (P = 0.068; P = 0.12), no statistically significant differences were seen among Sarasota Bay, Charlotte Harbor, and other published estimates. Although lacaziosis is a rare disease among these dolphin populations, studies that assess disease burden among different populations can assist with the surveillance of this zoonotic pathogen.     

Striped dolphins and short‐beaked common dolphins in the Gulf of Corinth,Greece: Abundance estimates from dorsal fin photographs

Capture‐recapture methods relying on dorsal fin natural markings have never been applied successfully to striped dolphins, Stenella coeruleoalba, and were rarely used to assess abundance of short‐beaked common dolphins, Delphinus delphis. We used digital photo‐identification to obtain abundance estimates of striped and common dolphins living in mixed groups in the Gulf of Corinth, Greece. The proportion of either species was calculated based on the relative number of photographs of adult animals showing relevant portions of their body during conspicuous surfacings. Striped dolphins and common dolphins averaged 95.0% and 3.2% of all individuals, respectively. Animals showing intermediate pigmentation accounted for another 1.8%. Striped dolphin numbers were relatively high, with a point estimate of 835 animals (95% CI = 631–1,106). Common dolphins numbers were low (point estimate 28 animals; 95% CI = 11–73) and individuals were scattered within striped dolphin groups, indicating that this common dolphin population may be nonviable. Within a semiclosed Gulf exposed to considerable anthropogenic impact, the future of both dolphin species is of concern due to their suspected geographic isolation and restricted extent of occurrence. Information provided here can be used to inform timely conservation efforts.     

ABUNDANCE OF IRRAWADDY DOLPHINS (ORCAELLA BREVIROSTRIS) AND GANGES RIVER DOLPHINS (PLATANISTA GANGETICA GANGETICA) ESTIMATED USING CONCURRENT COUNTS MADE BY INDEPENDENT TEAMS IN WATERWAYS OF THE SUNDARBANS MANGROVE FOREST IN BANGLADESH

Independent observer teams made concurrent counts of Irrawaddy dolphins Orcaella brevirostris and Ganges River dolphins Platanista gangetica gangetica in mangrove channels of the Sundarbans Delta in Bangladesh. These counts were corrected for missed groups using mark-recapture models. For Irrawaddy dolphins, a stratified Lincoln-Petersen model, which incorporated group size and sighting conditions as covariates, and a Huggins conditional likelihood model, which averaged models that individually incorporated group size, sighting conditions, and channel width as covariates, generated abundance estimates of 397 individuals (CV = 10.2%) and 451 individuals (CV = 9.6%), respectively. For Ganges River dolphins, a stratified Lincoln-Petersen model, which incorporated group size as a covariate, and a Huggins conditional likelihood model, which averaged the same models described above, generated abundance estimates of 196 individuals (CV = 12.7%) and 225 individuals (CV = 12.6%), respectively. Although the estimates for both models were relatively close, the analytical advantages of the Huggins models probably outweigh those of the Lincoln-Petersen models. However, the latter should be considered appropriate when simplicity is a priority. This study found that waterways of the Sundarbans support significant numbers of Irrawaddy and Ganges River dolphins, especially compared to other areas where the species have been surveyed.     

It counts who counts: an experimental evaluation of the importance of observer effects on spotlight count estimates

Spotlight surveys conducted by volunteers is a promising method to assess the abundance of nocturnally active mammals, but estimates are subject to bias if different observer groups differ in their ability to detect animals in the dark. We quantified the variation amongst volunteer spotlight observers with respect to their ability to detect and estimate distance to realistic animal silhouettes at different distances. Detection probabilities were higher for observers experienced in spotlighting mammals than for inexperienced observers, higher for observers with a hunting background compared with non-hunters and decreased as function of age but were independent of sex or educational background. If observer-specific detection probabilities were applied to real counting routes, point count estimates from inexperienced observers without a hunting background would only be 43 % (95 % CI, 39–48) of what inexperienced hunters with a hunting background would obtain and 29 % (25–33) of what experienced spotlight observers would detect. Mean estimated distances to objects did not deviate from true distances (no bias) but were highly imprecise. Female non-hunters estimated distances less precisely than other observers and precision increased with age. The study shows that observer effects may influence abundance estimates and underlines the importance of testing and accounting for observer effects when designing citizen science-based population survey programmes.     

A Double-Observer Method to Estimate Detection Rate During Aerial Waterfowl Surveys

Abstract We evaluated double-observer methods for aerial surveys as a means to adjust counts of waterfowl for incomplete detection. We conducted our study in eastern Canada and the northeast United States utilizing 3 aerial-survey crews flying 3 different types of fixed-wing aircraft. We reconciled counts of front- and rear-seat observers immediately following an observation by the rear-seat observer (i.e., on-the-fly reconciliation). We evaluated 6 a priori models containing a combination of several factors thought to influence detection probability including observer, seat position, aircraft type, and group size. We analyzed data for American black ducks (Anas rubripes) and mallards (A. platyrhynchos), which are among the most abundant duck species in this region. The best-supported model for both black ducks and mallards included observer effects. Sample sizes of black ducks were sufficient to estimate observer-specific detection rates for each crew. Estimated detection rates for black ducks were 0.62 (SE = 0.10), 0.63 (SE = 0.06), and 0.74 (SE = 0.07) for pilot-observers, 0.61 (SE = 0.08), 0.62 (SE = 0.06), and 0.81 (SE = 0.07) for other front-seat observers, and 0.43 (SE = 0.05), 0.58 (SE = 0.06), and 0.73 (SE = 0.04) for rear-seat observers. For mallards, sample sizes were adequate to generate stable maximum-likelihood estimates of observer-specific detection rates for only one aerial crew. Estimated observer-specific detection rates for that crew were 0.84 (SE = 0.04) for the pilot-observer, 0.74 (SE = 0.05) for the other front-seat observer, and 0.47 (SE = 0.03) for the rear-seat observer. Estimated observer detection rates were confounded by the position of the seat occupied by an observer, because observers did not switch seats, and by land-cover because vegetation and landform varied among crew areas. Double-observer methods with on-the-fly reconciliation, although not without challenges, offer one viable option to account for detection bias in aerial waterfowl surveys where birds are distributed at low density in remote areas that are inaccessible by ground crews. Double-observer methods, however, estimate only detection rate of animals that are potentially observable given the survey method applied. Auxiliary data and methods must be considered to estimate overall detection rate.     

Spatiotemporal patterns of marine mammal distribution in coastal waters of Galicia, NW Spain

The spatial and seasonal distribution of cetaceans and possible links with environmental conditions were studied at the Galician continental shelf. Data were collected between February–August 2001 and June–September 2003 during opportunistic surveys onboard fishing boats. Seven species of cetaceans were identified from 250 sightings of 6,846 individuals. The common dolphin (Delphinus delphis) was by far the most frequently sighted and the most widely distributed species. Spatiotemporal trends in cetacean distribution and abundance, and their relationships with environmental parameters (sea depth, SST and chlorophyll-a) were quantified using generalised additive models (GAMs). Results for all cetaceans were essentially the same as for common dolphins alone. Modelling results indicated that the number of common dolphin sightings per unit effort was higher further south. The number of individual common dolphins seen per sighting of this species (i.e. group size) was however higher in the north and west of the study area, higher later in the year and higher in 2001 than in 2003. In contrast, the number of common dolphin calves seen (per sighting of this species) was higher in the south. Models including environmental variables indicated larger common dolphin group sizes in deeper waters and at higher chlorophyll concentrations (i.e. in more productive areas). There was also a positive relationship between survey effort and group size, which is probably an artefact of the tendency of the survey platforms (fishing boats) to spend most time in areas of high fish abundance. Numbers of common dolphin calves per sighting were found to be higher in shallower waters. The results are consistent with common dolphins foraging mainly in deeper waters of the Galician continental shelf, while more southern inshore waters may represent a nursery area.     

A comparison of image and observer based aerial surveys of narwhal

From 25 to 30 August 2014 a double‐observer line‐transect survey was conducted over Melville Bay, home to one of two summering populations of narwhal (Monodon monoceros) off West Greenland. A total of 1,932 linear kilometers was surveyed along 33 transects. In addition to using observers, the aircraft was equipped with two oblique cameras to capture a comparable data set. Analysts reviewed the images for narwhal sightings, which were then matched to the observer sightings. The objectives of the study were to determine advantages and disadvantages of the detection capabilities of both methodologies, and to conduct a comparative analysis of population abundance estimates. Correcting for the truncated detection distance of the images (500 m), the image analysts recorded more sightings (62) and a lower mean group size (2.2) compared to aerial observers (36 and 3.5, respectively), resulting in comparable numbers of individuals detected by both platforms (135 vs. 126). The abundance estimate based on the image sightings was 2,536 (CV = 0.51, 95% CI: 1,003–6,406), which was not significantly different from the aerial observers estimate of 2,596 individuals (CV = 0.51; 95% CI: 961–7,008). This study supports the potential of using UAS for marine mammal abundance studies.     

DIFFERENCES BETWEEN FISHERY-DEPENDENT AND FISHERY-INDEPENDENT ESTIMATES OF SINGLE-AND MIXED-SPECIES DOLPHIN SCHOOLS: IMPLICATIONS FOR SINGLE-SPECIES STOCK ASSESSMENTS

Since 1979, fishery-independent data and fishery-dependent data have been used to estimate trends in the abundance of the northeastern stock of spotted dolphins in the eastern tropical Pacific. Data collected aboard tuna vessels have been used to estimate trends in relative abundance, while data collected from fishery-independent research vessels have been used to estimate trends in actual abundance. One of the largest discrepancies between the two data sources is that tuna vessels tend to report dolphin schools that are 400%—500% larger than schools observed by research vessels. After comparing research vessel and tuna vessel observations overlapping in space and time, it appears that either measurement error or selective reporting of large schools is the most likely explanation for the disparity. Comparing single-species schools against portions of mixed- species schools introduced bias. Revised fishery-dependent abundance estimates were generated using only single-species dolphin schools, resulting in estimates that appear more similar to recent fishery-independent abundance estimates.
correctable in fishery-dependent data, suggesting that data on mixed-species schools should not be considered if fishery-dependent data are to be combined with fishery-independent data for stock assessment purposes.     

To See or Not to See: Investigating Detectability of Ganges River Dolphins Using a Combined Visual-Acoustic Survey

Detection of animals during visual surveys is rarely perfect or constant, and failure to account for imperfect detectability affects the accuracy of abundance estimates. Freshwater cetaceans are among the most threatened group of mammals, and visual surveys are a commonly employed method for estimating population size despite concerns over imperfect and unquantified detectability. We used a combined visual-acoustic survey to estimate detectability of Ganges River dolphins (Platanista gangetica gangetica) in four waterways of southern Bangladesh. The combined visual-acoustic survey resulted in consistently higher detectability than a single observer-team visual survey, thereby improving power to detect trends. Visual detectability was particularly low for dolphins close to meanders where these habitat features temporarily block the view of the preceding river surface. This systematic bias in detectability during visual-only surveys may lead researchers to underestimate the importance of heavily meandering river reaches. Although the benefits of acoustic surveys are increasingly recognised for marine cetaceans, they have not been widely used for monitoring abundance of freshwater cetaceans due to perceived costs and technical skill requirements. We show that acoustic surveys are in fact a relatively cost-effective approach for surveying freshwater cetaceans, once it is acknowledged that methods that do not account for imperfect detectability are of limited value for monitoring.     

PRIMARY PRODUCTION REQUIRED TO SUPPORT BOTTLENOSE DOLPHINS IN A SALT MARSH ESTUARINE CREEK SYSTEM

We developed a model to estimate the proportion of annual primary production required to support bottlenose dolphins within the 32-km² North Inlet salt marsh creek system in South Carolina, U. S. A. The estimated annual prey consumption by dolphins was compared to the total annual production of prey available to dolphins, as determined from estimates of annual primary production, trophic transfer efficiencies, and the mean trophic level of prey. A best estimate range of 3.2%-6.8% of the total annual primary production of the North Inlet system was required to support an average population of only six dolphins (maximum range of 0.4%–7.0%). Dolphins were estimated to consume 11.1–14.2 metric tons of fish (wet weight) each year in North Inlet. The proportion of North Inlet primary production required to support dolphins increased dramatically during the winter months, when primary production declined but dolphin numbers remained similar. This period was marked by a decline in the abundance of available prey species and by a shift in the creek utilization patterns of dolphins. Despite the numerical scarcity of dolphins in the system, they appear to have a significant ecological impact and may be important predators of overwintering prey species.     

SHARK ATTACKS ON BOTTLENOSE DOLPHINS (TURSIOPS ADUNCUS) IN SHARK BAY, WESTERN AUSTRALIA: ATTACK RATE, BITE SCAR FREQUENCIES, AND ATTACK SEASONALITY

Shark predation may have been a central factor influencing the evolution of sociality in dolphins, as well as a determinant of dolphin habitat use and behavior. To understand the role of predation in driving interpopulation differences in behavior and sociality, it is important to quantify differences in predation risk among populations. This study describes the frequency of shark-inflicted scars and estimates the shark attack rate on bottlenose dolphins ( Tursiops aduncus ) in Shark Bay, Western Australia. Shark bite scars were found on 74.2% (95 of 128) of non-calves, and most of these scars were inflicted by tiger sharks ( Galeocerdo cuvier ). Although there were no differences among age/sex classes in the frequency of scarring, significantly more adult males than adult females bore multiple scars. The rate of unsuccessful shark attack was estimated to be between 11% and 13% of dolphins attacked each year. Large sharks (>3 m) were responsible for a disproportionate number of attacks. However, bites from small carcharhinid sharks on 6.2% of dolphins suggest that some of these small sharks may be dolphin ectoparasites. Both the scar frequencies and attack rate suggest that Shark Bay dolphins face a greater risk of predation than bottlenose dolphins in other locations.     

Why Do Dolphins Form Mixed‐Species Associations in the Azores?

Mixed‐species associations are temporary associations between individuals of different species that are often observed in birds, primates and cetaceans. They have been interpreted as a strategy to reduce predation risk, enhance foraging success and/or provide a social advantage. In the archipelago of the Azores, four species of dolphins are commonly involved in mixed‐species associations: the common dolphin, Delphinus delphis, the bottlenose dolphin, Tursiops truncatus, the striped dolphin, Stenella coeruleoalba, and the spotted dolphin, Stenella frontalis. In order to understand the reasons why dolphins associate, we analysed field data collected since 1999 by research scientists and trained observers placed onboard fishing vessels. In total, 113 mixed‐species groups were observed out of 5720 sightings. The temporal distribution, habitat (water depth, distance to the coast), behaviour (i.e. feeding, travelling, socializing), size and composition of mixed‐species groups were compared with those of single‐species groups. Results did not support the predation avoidance hypothesis and gave little support to the social advantage hypothesis. The foraging advantage hypothesis was the most convincing. However, the benefits of mixed‐species associations appeared to depend on the species. Associations were likely to be opportunistic in the larger bottlenose dolphin, while there seemed to be some evolutionary constraints favouring associations in the rarer striped dolphin. Comparison with previous studies suggests that the formation of mixed‐species groups depends on several environmental factors, and therefore may constitute an adaptive response.     

Abundance and residency dynamics of the Indo-Pacific humpback dolphin,Sousa chinensis,in the Dafengjiang River Estuary,China

Robust population size estimates are essential for informing population conservation status. Residency dynamics show population habitat use through time. Population size of Indo-Pacific humpback dolphins (Sousa chinensis) has been extensively investigated in Chinese waters, but their residency dynamics are rarely known. Mark-recapture analysis based on photo-identification records was applied to humpback dolphins in the Dafengjiang River Estuary habitat, one of the key habitats in the northern Beibu Gulf, China. Movement analyses based on lagged identification rate indicated the humpback dolphins spent, on average, 78.5 days inside and 46.9 days outside the survey area. Within the study area, the humpback dolphin abundance was 83 identifiable dolphins. A total of 353–430 humpback dolphins, estimated by POPAN modeling, were involved in this fluid habitat-use dynamic. Robust Design analysis showed strong seasonality in humpback dolphin abundance and emigration probability, implying a movement- and habitat-use pattern likely associated with spatiotemporal distribution of oceanographic characteristics and prey occurrences. Population surveys and conservation measures currently conducted in Chinese waters seldom consider seasonality in movements between habitat patches, which can be addressed by genetic analyses across habitats and cross-matching photo-identification records among neighboring habitats.     

Observer error in vegetation surveys: a review

Aims Vegetation sampling employing observers is prone to both inter-observer and intra-observer error. Three types of errors are common: (i) overlooking error (i.e. not observing species actually present), (ii) misidentification error (i.e. not correctly identifying species) and (iii) estimation error (i.e. not accurately estimating abundance). I conducted a literature review of 59 articles that provided quantitative estimates or statistical inferences regarding observer error in vegetation studies.Important findings Almost all studies (92%) that tested for a statistically significant effect of observer error found at least one significant comparison. In surveys of species composition, mean pseudoturnover (the percentage of species overlooked by one observer but not another) was 10–30%. Species misidentification rates were on the order of 5–10%. The mean coefficient of variation (CV) among observers in surveys of vegetation cover was often several hundred % for species with low cover, although CVs of 25–50% were more representative of species with mean covers of>50%. A variety of metrics and indices (including commonly used diversity indices) and multivariate data analysis techniques (including ordinations and classifications) were found to be sensitive to observer error. Sources of error commonly include both characteristics of the vegetation (e.g. small size of populations, rarity, morphology, phenology) and attributes of the observers (e.g. mental fatigue, personal biases, differences in experience, physical stress). The use of multiple observers, additional training including active feedback approaches, and continual evaluation and calibration among observers are recommended as strategies to reduce observer error in vegetation surveys.